Hey Ryan,

if you add a DRAM interface to your YAML and call rfnoc_modtool add, then
you will get an interface. In the other thread that Brian mentions, he and
Wade discuss that it might be too wide -- but that should not be a problem.

Give it a shot, and let us know if you run into any issues.

--M

On Wed, Oct 8, 2025 at 8:50 PM Brian Padalino <[email protected]> wrote:

> On Wed, Oct 8, 2025 at 2:40 PM <[email protected]> wrote:
>
>> Greetings,
>>
>> I'm writing to request a heading-check on the design of an RFNoC block
>> that needs to conditionally save/recall an arbitrary number of samples. I
>> successfully built & ran a version of the gain example to convince myself
>> that the build environment was setup correctly. Now I'm struggling to add
>> the DDR interface to the design.
>>
>> My first thought was to port the functionality of the replay block to an
>> OOT block that contains decision logic to save/recall the samples. By using
>> the replay block's YAML file as a starting point, I expected rfnoc_modtool
>> to generate HDL similar to what can be seen in rfnoc_block_replay.v.
>> Unfortunately, the AXI memory mapped interface is absent in the resulting
>> HDL. That makes me think that my YAML definition lacks specificity, or that
>> the interface was added manually. I understand rfnoc_modtool is meant to be
>> a starting point, so am I expecting too much?
>>
> I believe I ran into a similar issue when I was asking about it here:
>
>
> https://lists.ettus.com/empathy/thread/6UOYGNTMTCP2SFP4PKYH2RF3IWPXWQVJ?hash=6UOYGNTMTCP2SFP4PKYH2RF3IWPXWQVJ#6UOYGNTMTCP2SFP4PKYH2RF3IWPXWQVJ
>
> I solved it by hand writing my HDL since I am more proficient in that, but
> the links to the files in the above post are the places to modify the code
> generation to get the HDL to better reflect the interfaces you're
> requesting.
>
> HDL is not in my wheelhouse, but I'm working with another engineer who has
>> experience. We appreciate any insight & advice that this forum provides.
>>
>
> Good luck - hopefully a pull request can come from this work since it's a
> useful interface for blocks.
>
> Brian
> _______________________________________________
> USRP-users mailing list -- [email protected]
> To unsubscribe send an email to [email protected]
>
_______________________________________________
USRP-users mailing list -- [email protected]
To unsubscribe send an email to [email protected]

Reply via email to